148 Spirituality Studies 3 Results 3.1 Prerequisite Tests of Analysis 3.1.1 Prerequisite Tests of Analysis Prior to conducting inferential statistical analyses, prerequisite tests were performed to ensure that the data met the assumptions required for parametric procedures. Normality and homogeneity of covariance matrices were tested for both dimensions simultaneously and are presented in Table 4. For normality, assessed using the One-Sample Kolmogorov–Smirnov Test, all Asymp. Sig. values exceeded the .05 threshold for both SOM (pre-test p =.052, post-test p =.062, follow-up p =.054) and OOM (pre-test p =.062, post-test p =.065, follow-up p =.052), confirming that the normality assumption was satisfied for both dimensions at all time points. For homogeneity of covariance matrices, Box’s M test indicated no significant violation for SOM (Box’s M = 7.762, F = 1.205, p =.300) or OOM (Box’s M = 10.932, F = 1.697, p =.117), supporting the use of multivariate analyses. 3.2 Descriptive Statistics Descriptive statistics for both SOM and OOM dimensions by gender and measurement point are presented in Table 5. Male participants consistently reported higher mean scores than female participants across all time points for both dimensions. For SOM, total scores showed a gradual decline across measurement points, observable in both male and female participants a negative finding discussed further in the Discussion section. For OOM, total scores remained relatively stable, with male scores increasing and female scores declining over time. Given the pre-experimental design without a control group, these descriptive patterns are interpreted cautiously and do not imply causal effects of the intervention. 3.3 Multivariate Tests Multivariate analyses were conducted to examine gender-related differences across both meaning dimensions simultaneously. The combined results for SOM and OOM are presented in Table 6. The between-subjects effect of gender was statistically significant across both dimensions (Wilks’ λ =.806, F(2,47) = 5.648, p =.006, partial η² =.194), indicating a large effect size. The main effect of Time was not statistically significant (Wilks’ λ =.836, F(4,45) = 2.215, p =.082, partial η² =.164), suggesting that overall change across measurement points was not significant independent of gender. Notably, the Time × Gender interaction was statistically significant and large in practical magnitude (Wilks’ λ =.717, F(4,45) = 4.443, p =.004, partial η² =.283), indicating that the pattern of change over time differed meaningfully between male and female participants. 3.4 Between-Subjects Effects Between-subjects effects were examined to assess gender differences at each measurement point and overall. Results are presented in Table 7. The overall effect of gender was significant and large for both dimensions: SOM (F(1,48) = 10.728, p =.002, partial η² =.183) and OOM (F(1,48) = 11.528, p =.001, partial η² =.194). Gender differences were also significant at each individual time point for both dimensions, indicating that gender-related differences were stable and consistent across the full observation period rather than emerging only at specific time points. 3.5 Pairwise Comparisons Pairwise comparisons were conducted to clarify the direction and magnitude of gender differences across both meaning dimensions, with results presented in Table 8. Male participants scored significantly higher than female participants on both SOM and OOM, with large effect sizes for both dimensions. Confidence intervals for both comparisons excluded zero, confirming the statistical reliability of the gender gap. These results confirm H1 and H2 respectively. 3.6 Estimated Marginal Means Estimated marginal means by gender and time point for both dimensions are presented in Table 9. Male participants consistently reported higher scores than female participants across all time points for both SOM and OOM. For SOM, both male and female participants showed a gradual decline from pre-test to follow-up. For OOM, male scores showed a gradual increase over time while female scores showed a gradual decline, reflecting the significant Time × Gender interaction reported in Table 4.
RkJQdWJsaXNoZXIy MTUwMDU5Ng==